The package bundles data named txhousing which contains monthly housing sales in various Texan cities.

library(plotly)
if (interactive()) View(txhousing) else txhousing
## # A tibble: 8,602 × 9
##       city  year month sales   volume median listings inventory     date
##      <chr> <int> <int> <dbl>    <dbl>  <dbl>    <dbl>     <dbl>    <dbl>
## 1  Abilene  2000     1    72  5380000  71400      701       6.3 2000.000
## 2  Abilene  2000     2    98  6505000  58700      746       6.6 2000.083
## 3  Abilene  2000     3   130  9285000  58100      784       6.8 2000.167
## 4  Abilene  2000     4    98  9730000  68600      785       6.9 2000.250
## 5  Abilene  2000     5   141 10590000  67300      794       6.8 2000.333
## 6  Abilene  2000     6   156 13910000  66900      780       6.6 2000.417
## 7  Abilene  2000     7   152 12635000  73500      742       6.2 2000.500
## 8  Abilene  2000     8   131 10710000  75000      765       6.4 2000.583
## 9  Abilene  2000     9   104  7615000  64500      771       6.5 2000.667
## 10 Abilene  2000    10   101  7040000  59300      764       6.6 2000.750
## # ... with 8,592 more rows

The ggplotly() function converts a ggplot object to a plotly object. If you already use ggplot2, this makes it easy to leverage simple interactive techniques (tooltips, pan, and zoom) .

p <- ggplot(txhousing) +
  geom_line(aes(x = date, y = median, group = city))

ggplotly(p)

As a side-note, using %>% (aka, the pipe operator) allows humans to read your code from left-to-right rather than inside-out. It does this by inserting the object on left-hand side into a function on the right-hand side.

# this code is equivalent to ggplotly(p)
p %>% ggplotly()

The R package also has a non-ggplot2 interface which embraces the pipe and works well with dplyr’s data manipulation verbs (a la ggvis). To read more about this interface, see this section of the plotly for R book.

txhousing %>%
  group_by(city) %>%
  plot_ly(x = ~date, y = ~median) %>%
  add_lines(name = "All cities") %>%
  filter(city == "Houston") %>%
  add_lines(name = "Just Houston")

Highlighting (or annotating) something interesting is generally a good idea when presenting graphics to someone else. However, we often don’t know what is actually interesting at runtime. In this case, it is nice to have tools for interactively highlighting interesting “units”. The SharedData class from the crosstalk package provides a mechanism for you to declare an “interaction unit”. In this case, we specify city as the interaction unit, then use the highlight() function from plotly for extra control over the type of interaction which triggers a selection.

library(crosstalk)
tx <- SharedData$new(txhousing, ~city, "Selected cities")

p <- ggplot(tx) +
  geom_line(aes(x = date, y = median, group = city))

p %>%
  ggplotly(tooltip = "city") %>%
  highlight(
    on = "plotly_hover", 
    off = "plotly_doubleclick", 
    dynamic = TRUE,
    persistent = TRUE,
    selectize = TRUE
  )

Your Turn: Use the color argument to supply your own dynamic color palette. Highlight ‘South Padre Island’ one color and ‘San Marcos’ another. What do you see? Is there a better way we could ‘guide’ selections if we are interested in understanding missing values? See here for my solution.

Highlighting can help combat overplotting, but becomes much more powerful when applied to multiple linked views. There are a number of ways to arrange and link multiple views – one of them being via ggplot2’s facetting (i.e., trellis display) functionality. A trellis display is a special class of “small multiples” where each panel displays a different subset of the data. Below is a trellis display of median house price versus of month of year across different cities. Since there are multiple years for each city, we can set the “interaction unit” to year and explore how house prices fluctuate across cities for a given year.

tx <- SharedData$new(txhousing, ~year, "Selected years")

p <- ggplot(tx) +
  geom_line(aes(x = month, y = median, group = year)) + 
  facet_wrap(~city)

p %>%
  ggplotly(tooltip = "year", width = 1200, height = 600) %>%
  highlight(
    on = "plotly_click", 
    off = "plotly_doubleclick", 
    selectize = TRUE
  )

Your Turn: ggplotly() can also convert ggmatrix objects from the GGally package (ggmatrix objects are the foundation of the ggpairs() and ggduo() functions). First, make a scatterplot matrix of median, volume, listings, and inventory via ggpairs(). Second, make it interactive with ggplotly() and highlight points with high inventory. What do you see?

Sometimes it’s useful to “animate selections” – especially if the unit of interest has a natural ordering. Both ggplotly() and plot_ly() respect a frame aesthetic, which is similar to facetting, but panels are overlayed in an animation, rather than into small multiples.

txhousing %>%
  filter(city == "Houston") %>%
  group_by(year) %>%
  plot_ly(x = ~month, y = ~median, color = I("black")) %>%
  add_lines(alpha = 0.2) %>%
  add_lines(frame = ~year, ids = ~month) %>%
  hide_legend()

Your turn: 1. Use the animation_opts() function to alter the speed and easing of the animation above. 2. Modify our first faceted plot to animate through the years 

d <- SharedData$new(mpg)
dots <- plot_ly(d, color = ~class, x = ~displ, y = ~cyl)
boxs <- plot_ly(d, color = ~class, x = ~class, y = ~cty) %>% add_boxplot()
bars <- plot_ly(d, x = ~class, color = ~class)

subplot(dots, boxs) %>%
  subplot(bars, nrows = 2) %>%
  layout(
    dragmode = "select",
    barmode = "overlay",
    showlegend = FALSE
  )